In the complex and ever-changing urban gas application environment, traditional detection methods suffer from significant shortcomings in sensitivity, mobility, and real-time performance, thereby limiting the requirements for精细化巡检. Consequently, this study focuses on gas leak detection tasks, analyzes the selection and integration needs of high-sensitivity detection equipment, and constructs an integrated system architecture based on a robotic dog platform. Modular design and implementation have been carried out for the structural, power, control, and communication subsystems, effectively enhancing the deployment capability and detection performance of the equipment in real-world environments.